![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
School of Biological Sciences, Faculty of Science (J.L., Y.-K.C., A.A.O., G.J.S.C.), Department of Medicine, Faculty of Medical and Health Sciences (G.D.G., G.J.S.C.), Human Nutrition Unit (S.D.P.), and Centre for Molecular Biodiscovery (G.J.S.C.), University of Auckland, Auckland, New Zealand
(Received September 14, 2006; Accepted November 8, 2006)
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
; Struthers and Morris, 2002). Recently, we showed that triethylenetetramine (TETA), a CuII-selective chelator used for the treatment of Wilson's disease (Walshe, 1982), ameliorated left ventricular hypertrophy in diabetic rats and humans (Cooper et al., 2004). We proposed that metabolic changes in diabetes, in particular, those resulting in the accumulation of advanced glycation endproducts in the extracellular matrix, lead to or cause the accumulation of excess extracellular CuII in the extracellular matrix, and that the resulting tissue-Cu imbalance could cause heart failure (Cooper et al., 2004, 2005). The ability of TETA to remove excess extracellular CuII thus makes it a candidate pharmacotherapeutic for the cardiovascular complications of diabetes. Consequently, it is now undergoing phase II clinical trials to assess its safety and efficacy as a therapy for heart failure in diabetes (Cooper et al., 2004).
Although TETA administration can elevate urinary Cu excretion (Cooper et al., 2005), the exact mechanism of its action and pharmacology in diabetes is still under investigation. There are relatively few reports on TETA pharmacology in patients with Wilson's disease in the literature, although it has been used as a treatment for this condition for decades (Walshe, 1982; Leggio et al., 2005). Indeed, most relevant human-derived pharmacological data have been derived from studies in healthy volunteers. Most oral TETA is not absorbed, but excreted unchanged in the feces (Gibbs and Walshe, 1986). The 5 to 18% of TETA that is systemically absorbed is said to be extensively metabolized, with the majority being excreted in urine as metabolite(s) (Kodama et al., 1993, 1997), and results from rat studies have been consistent (Kobayashi et al., 1990; Takeda et al., 1995a,b,c). One major metabolite, MAT, has been reported in human urine (Kodama et al., 1997). In those reports, TETA was fluorescence-derivatized and detected using HPLC (Nakano et al., 2002). Fluorescence derivatization methods are often associated with problems such as 1) whether the analyte is fully or partially labeled, 2) whether metabolites are labeled, 3) whether metabolites are fully or partially labeled, and 4) whether the analyte-specific signal is separated from those corresponding to other known or unknown metabolites.
Here, we used LC-MS to study TETA metabolism and excretion by analyzing urine samples from drug-treated diabetic subjects and healthy volunteers. We have detected a novel metabolite, N1,N10-diacetyltriethylenetetramine (DAT) and developed an LC-MS method to measure TETA and both its major metabolites in urine, within one injected sample and without fluorescence labeling. In addition, we concomitantly measured urinary concentrations of TETA and its major metabolites, and analyzed TETA-evoked changes in Cu and Zn excretion in relation to drug and metabolite concentrations.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
). All protocols received appropriate regulatory approval, and all subjects provided written informed consent. Authentic Standards. TETA dihydrochloride, MAT trihydrochloride, and DAT dihydrochloride, all of >99.9% purity, were synthesized and supplied by CarboGen AG (Hunzenschwil, Switzerland). N1-Acetylspermine (ACS) dihydrochloride, of greater than 98% purity, was obtained from Sigma-Aldrich (St. Louis, MO).
Collection of Urine and Plasma Samples. Subjects received increasing doses of TETA dihydrochloride (300, 600, 1200, and 2400 mg/day; Anstead, Essex, UK) in an open-label, dose-escalating trial. Each dose was taken at a single time in the morning (as 300-mg capsules) after an overnight fast, for a period of 7 days, with 6-week washout periods between each dose. Dosing was not witnessed; therefore, the possibility of incomplete compliance was addressed by drug capsule/p-aminobenzoic acid (PABA) tablet counts and determination of PABA recovery. For each subject, 24-h urine samples were collected immediately before and on the 7th day of each dosing schedule. Urine samples were collected in precleaned, wide-mouth, trace metal-free plastic receptacles (Nuplex Industries Ltd., Auckland, New Zealand). Gross urine volumes from each 24-h collection were recorded and urine collections were monitored using a PABA test (Cooper et al., 2005). Subjects were given PABA-check tablets (Laboratory for Applied Biology, London, UK) containing 80 mg of the naturally occurring B-vitamin complex to be taken at each of three main meals throughout the day. A recovery of >75% of the oral dose is indicative of complete urine collection over the 24-h period. Low recoveries provide a qualitative indicator of poor urine collection, although failure to adhere to the 80-mg three times per day dosing regimen cannot be excluded when recovery is low, and recovery of <5% is strongly indicative of complete noncompliance. The PABA results were used together with subjects' feedback on drug capsule and PABA tablet intake, completeness of urine collection, urine volume consistency, drug levels in urine, and copper and zinc changes in urine, to determine whether a data point was included in the final statistical analysis. Undiluted urine samples were transferred to 5-ml screw-cap tubes and stored at –80°C until analysis.
In a separate study, 2-ml heparinized blood samples were taken from one healthy male volunteer before and at 10, 20, 40, 60, 90, and 120 min after taking 2400 mg of oral TETA dihydrochloride. Each sample was centrifuged at 10,000g and plasma was collected into 1.5-ml tubes. Samples were stored at –80°C until analysis.
Urinary Cu and Zn Analysis. [Cu] and [Zn] were measured in pre- and post-treatment urine samples by inductively coupled plasma mass spectrometry (ICPMS; Elan 6100; PerkinElmerSciex Instruments, Boston, MA) as described previously (Cooper et al., 2005). In brief, diluted samples and matrix-matched standard curve solutions were measured by ICPMS with gallium as the internal standard. Cu and Zn concentrations were calculated from the standard curve. The ICPMS operating parameters were as follows: radiofrequency power, 1500 W; nebulizer gas flow rate, 0.9 l/min; auxiliary gas flow rate, 1.2 l/min; plasma gas flow rate, 15 l/min; reaction gas, NH3 at 0.3 ml/min; data acquisition mode, peak hopping, three replicates, 20 sweeps per replicate; and sample uptake rate, 1 ml/min.
Metabolite Identification. Urine samples (0.5 ml) were thawed, centrifuged, and transferred into 1.5-ml snap-cap metal-free HPLC vials. One hundred microliters of plasma was added to 50 µl of 10% heptafluorobutyric acid (HFBA; Sigma-Aldrich) and centrifuged to remove precipitated protein. Urine and plasma supernatants were first injected into the mass spectrometer directly. MS conditions were optimized as follows: curved desolvation line temperature, 200°C; heat block, 200°C; nebulizing gas, 1.5 l of N2/min; interface voltage, 3.5 kV; CDL voltage, 40.0 V; and Q-array voltage, DC 0.0 V, radiofrequency 140 V. The MS positive-ion scan mode was used and ion profiles from pre- and postdrug samples were compared to detect putative major metabolites. Supernatant was then transferred into 1.5-ml HPLC vials, which were loaded onto the temperature-controlled auto-injector of an LC-MS system, comprising an LC-10ADvp pump, a DGU-14AM degasser, a SIL-10ADvp temperature-controlled auto-injector, a CTO-10ASvp column oven, and a 2010A single quadrupole mass spectrometer equipped with an electrical spray ionization interface (Shimadzu Corp., Kyoto, Japan). Fifty-microliter samples were injected into the LC-MS apparatus and chromatographic separation was achieved using a 5-µ Cyano 100 x 4.6 mm column (maintained at 25°C in the column oven) with guard (Phenomenex, Torrance, CA) using 15% (v/v) acetonitrile, 0.1% (v/v) heptafluorobutyric acid in Milli Q water (18 M
resistivity; Waters, Milford, MA) as mobile phase. The flow rate was 0.5 ml/min. All samples were analyzed first using MS positive-ion scan mode, with molecular mass range from 50 to 1000 Da, to identify peaks corresponding to TETA and its metabolite(s). Samples were then reanalyzed together with blank urine, to which had been added the authentic standards (TETA, MAT, and DAT), by LC-MS with an additional MS-positive selected-ion monitoring (SIM) signal set at [M + H]+ = 147, 189, and 231 Da, with a micro-scan of 0.5 Da (corresponding to TETA and the metabolites found in previous MS scans), to better identify and characterize the analytes.
Measurement of Urinary Metabolites. Serial dilutions of TETA, MAT, and DAT standards were made in blank urine samples to create standard curves, such that each sample contained one aliquot of each standard. Twenty microliters of HFBA (600 mM) was added into 300 µl of each standard urine sample, and to urine samples from human subjects (with appropriate dilution as necessary). Samples were centrifuged (12,000g, 10 min, 4°C), and 200 µl of supernatant was then transferred to a HPLC vial, together with 50 µl of 100 µM ACS as internal standard. Final concentrations of HFBA and ACS were 30 mM and 20 µM, respectively. HFBA (30 mM) stabilized TETA in solution for at least 24 h. ACS appeared to be a suitable internal standard inasmuch as urinary excretion of natural endogenous ACS was undetectable in our experiments. Duplicate 50-µl sample aliquots were analyzed by LC-MS. Separation and detection of TETA and its metabolites were achieved under the same conditions as described above for metabolite identification, with the modification that an extra molecular mass, of 245 Da, was added to the SIM signal to detect the internal standard, ACS. All standards and samples were analyzed within 24 h of preparation.
Data acquisition and MS peak area integration for TETA, MAT, DAT, and ACS peaks were achieved using Shimadzu LCMSsolution software. To construct calibration curves, the area ratios of MS peaks for TETA or its metabolites, relative to that of the internal standard, were plotted against the standard concentrations. Lines of best fit were plotted by nonlinear hyperbolic regression analysis (Prism; GraphPad Software Inc., San Diego, CA). The range of the calibration curves was 1 to 16 µM for each standard (r2 = 0.99 in each case). The relative recoveries were 90 to 120% and coefficient of variation was <10%. The back-calculated values were within 10% of prepared values in each case.
|
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
). Another study measuring 0- to 26-h TETA and MAT in urine of healthy volunteers showed that TETA was not detectable after 10 h, and less than 3% of 0- to 26-h MAT was found in urine between 24 and 26 h (Kodama et al., 1997). Concentrations of TETA and its metabolites in 24-h urine samples were determined by comparison with standard curves (Fig. 4), and 24-h urinary excretion rates were calculated. Total 24-h urinary drug excretion (of all forms of the drug) was calculated as the sum of the amount of TETA and metabolites, and results are summarized in Tables 1 (for healthy volunteers) and 2 (for diabetic patients). Since all incidences of detection under the lower limit of quantification (LLOQ) for TETA and its metabolites occurred in healthy volunteers, the value of LLOQ (1 µM) was used to give a maximum estimate in subsequent statistical analysis. Since some subjects demonstrated apparent noncompliance by the criteria of PABA and/or capsule counts, a combination of PABA results, drug levels, Cu and Zn changes, urine volume consistency, and subjects' feedback of PABA and drug intake were used as inclusion/exclusion criteria to select data points for the final analysis (Tables 1 and 2; excluded data points are italicized). For example, if a subject did not take PABA (i.e., patient Dia5 with a PABA recovery of 2.5% at the 2400-mg dose) but still took the drug and completed urine collection (verified by urine volume consistency), the data point was included. On the other hand, if a subject had an acceptable PABA recovery (i.e., healthy subject HV6 with a PABA recovery of 79.4% at 300 mg dose) but did not take the drug, the data point was excluded.
|
|
|
As statistical models, LMEs were used to determine the significance of differences using all data points from different doses, because this technique can take multiple factors (dose, disease status, and drug concentration) into consideration simultaneously. There was no significant difference between the amount of unchanged TETA excreted in urine of healthy volunteers and diabetic patients (Fig. 5A; p = 0.873, LME, S-Plus). However, diabetic patients excreted significantly more metabolized TETA than did healthy volunteers (Fig. 5B; p = 0.015, LME). Total drug (including metabolites) excreted in urine was 0.03 to 13.4% (range) (3.2 ± 1.1%, mean ± S.E.M.) of administered dose in healthy volunteers and 3.7 to 14.6% (9.2 ± 0.4% %) of administered dose in diabetic patients. Thus, diabetic patients excreted significantly higher percentages of administered doses than healthy volunteers (Fig. 6; p = 0.005, LME), whereas dose did not affect the percentage of administered drug excreted in urine (p = 0.121, LME). In both diabetic and healthy subjects, there was a significant correlation between dose and amount of drug excreted (p < 0.001, LME), but there was no combined effect of dose and disease status on drug excretion (both parent and metabolites) (p > 0.05, LME). Within urinary drug (parent + metabolites), diabetic subjects demonstrated higher percentages of metabolites (92.9 ± 0.1%) than did healthy volunteers (77.2 ± 0.2%) (p = 0.007, LME), but dose elevation did not alter this difference (p = 0.224, LME).
|
|
Human 24-h Urinary Excretion of TETA and Its Metabolites, and Their Correlation with Changes in Urinary Cu and Zn Excretion. The changes in 24-h Cu and Zn excretion are as shown in Tables 1 and 2. They were calculated as 
Cu or Zn = [Cu or Zn content of 24-h urine after dosing] – [Cu or Zn content in 24-h urine before dosing]. Cu and Zn were closely related to dose elevation (p < 0.001, LME). Changes in Cu and Zn excretion in healthy volunteers and diabetic patients were then analyzed against amounts of urinary TETA, MAT, DAT, [TETA + MAT], or [TETA plus both metabolites].
TETA and MAT were both found to be significant independent predictors of urinary Cu excretion over drug levels for healthy volunteers and diabetic patients (Table 3). Further refinement of the model shown in Table 4 suggested that TETA alone is the best predictor of change in Cu in response to drug in healthy volunteers, whereas the sum of TETA and MAT is the best predictor for diabetic patients. For Zn, TETA alone was the best predictor of change in urinary Zn response to drug in both healthy and diabetic subjects (Table 3).
|
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
), whereas, to our knowledge, the existence of the second, DAT, has not previously been reported. Identifying major metabolites of TETA is an important step for further pharmacological investigation, such as absorption, distribution, metabolism and excretion (ADME) studies, in diabetic patients.
Previous studies showed that TETA was poorly absorbed. A study using [14C]TETA showed that only 6 to 18% of orally administered drug was systemically absorbed (Gibbs and Walshe, 1986). Another rat study reported the bioavailability of TETA to be about 13.8% (Tanabe, 1996). In an additional report, the urinary recovery of drug was about 9% of administered dose in healthy humans (Kodama et al., 1997). Our current results showed that urinary recoveries of drug varied between 0.03 and 13.4% of administered dose in healthy volunteers and between 3.7 and 14.6% in diabetic subjects, consistent with these previous studies. Our direct injection studies did not detect any other significant TETA metabolites in urine or plasma. Those findings, together with our recovery levels that were consistent with the previous [14C]TETA study, indicate that no other major metabolites of TETA are likely to be excreted in urine.
Most urinary TETA was present as metabolites, and unchanged TETA only accounted for 22.8 ± 0.6% and 7.1 ± 0.8% of total excreted drug in healthy and diabetic subjects, respectively. In drug excreted via the urine, the percentage of metabolized drug was much higher in diabetic subjects (92.9 ± 0.1%) than in healthy controls (77.2 ± 0.2%). This is in accord with previous findings that more metabolite than parent compound is excreted in urine of humans and rats (Kobayashi et al., 1990; Kodama et al., 1993, 1997; Takeda et al., 1995b). The greater amount of metabolized TETA excreted in diabetic patients is consistent with the hypothesis that TETA might be more extensively metabolized in them than in healthy controls (Fig. 5B). In a recent pharmacokinetic study, we found that MAT and DAT concentrations in the blood of diabetic patients were higher than those in healthy subjects at the same dose (600 mg). The Tmax values corresponding to TETA, MAT, and DAT in healthy subjects (n = 24) were 1.9 ± 0.8, 5.1 ± 0.6, and 5.1 ± 0.7 h, respectively. Serum Cmax values for TETA, MAT, and DAT in healthy subjects were 5.4 ± 2.3, 4.0 ± 1.6, and 1.1 ± 0.6 µM, respectively (unpublished data). In diabetic subjects, concentrations of TETA, MAT, and DAT were determined in serum sampled 4 to 6 h after dosing, and were 6.3 ± 4.2, 10.5 ± 5.3, and 2.6 ± 1.6 µM, respectively (unpublished data). Although there were no statistically significant differences between TETA concentrations, those of MAT and DAT in diabetic subjects were significantly higher than corresponding values in healthy subjects. This observation further supports our hypothesis that TETA might be metabolized faster in diabetic subjects than in healthy subjects.
One possibility is that diabetic patients may have higher hepatic metabolic rates for TETA. For example, previous rat (Sheweita et al., 2002) and rabbit (Arinc et al., 2005) studies indicated that diabetes increases the activity of certain hepatic enzymes (e.g., those in the cytochrome P450 family). There is also a report that type 2 diabetic patients had a marked increase in hepatic CYP2E1 activity (Wang et al., 2003). However, the nature of the enzymes responsible for TETA metabolism [presumably either N-acetyltransferase (NAT) or polyamine acetyltransferase, or both], and whether these enzyme activities are modified by diabetes, are still under investigation. NAT is an enzyme that could mediate TETA metabolism. Polymorphism of NAT could greatly affect NAT expression and, thus, TETA metabolism (Wormhoudt et al., 1999). We have recently studied the plasma pharmacokinetics in two healthy volunteer groups, with fast and slow acetylator phenotypes, and found no difference in any of the pharmacokinetic parameters between the two groups (unpublished data). However, whether diabetic subjects have higher rates of NAT mutation than normal subjects is unknown. We have planned a study to genotype and phenotype NAT in normal and diabetic subjects to determine whether there is any difference. The expression of metabolizing enzymes also changes with age. However, subjects in our study were age-matched; therefore, age was unlikely to influence these results. Another possibility is that, apart from the liver, TETA may be metabolized more extensively by enzyme(s) in other organs such as heart, kidney, or gut, in diabetic than in healthy people. We are currently investigating TETA metabolism in liver, kidney, and heart of healthy and diabetic rats to test these hypotheses.
The evidence that diabetic patients metabolize TETA more completely than healthy people suggests that further ADME studies in diabetic patients may be required. Previous ADME results from healthy volunteers for studies of Wilson's disease may not be applicable in diabetic patients, because a higher metabolic rate of TETA may change its ADME behavior, as well as its pharmacokinetics and pharmacodynamics. Here, a significantly higher proportion of administered drug was excreted by diabetic patients than by healthy subjects (Fig. 6). There is, thus, a correlation between the apparently greater rate of TETA metabolism and its higher intake or bioavailability in diabetic subjects.
The Cu excretion changes in healthy volunteers mainly correlated with urinary [TETA] (Table 4): urinary [Cu] increased in proportion to urinary [TETA], consistent with a prior report (Kodama et al., 1997). Urinary [Zn] showed similar trends in healthy volunteers, increasing with urinary [TETA] (Table 4). This is not surprising, because TETA is a more potent chelator than its N-acetyl metabolite in vitro (Kodama et al., 1997).
Interestingly, increases in Cu excretion in diabetic patients showed a different correlation with urinary drug levels, in that their Cu values were more closely related to [TETA + MAT] (Table 3) than for control subjects, in whom [TETA] alone was the main correlate. These findings are consistent with MAT acting as a Cu-chelator in vivo. Although MAT reportedly has in vitro Cu-chelating activity (Kodama et al., 1997), we know of no prior reports concerning its in vivo Cu- or Zn-chelating properties.
Our prior studies have shown that the mechanisms by which TETA elevates urinary Cu and Zn are likely to differ between diabetic and healthy people. TETA-mediated increases in urinary Cu and Zn in diabetic subjects resulted largely from their extraction from the systemic compartment whereas, by contrast, elevated urinary Cu and Zn excretion in healthy subjects resulted mainly from drug-mediated increases in metal uptake from the gut (Cooper et al., 2005). The difference between diabetic and control subjects may be due to the up-regulation of metallothionein in small bowel enterocytes in the former group, which could create a defensive barrier to the excessive Cu and Zn uptake that might otherwise result from diabetic hyperphagia.
In summary, we found two major TETA metabolites in human plasma and urine, one of which, DAT, has not previously been reported. Results of urinary drug analyses indicate that diabetic patients metabolize TETA more extensively than do healthy people, which in turn was associated with its higher uptake or bioavailability in those with the disease. These findings may warrant further ADME studies in diabetic patients. Increased urinary Zn excretion was mainly associated with unchanged urinary TETA in both healthy and diabetic subjects. Increased urinary Cu excretion in healthy volunteers showed similar characteristics, being mainly associated with unchanged urinary TETA levels. However, increased urinary Cu excretion in diabetic patients more closely correlated with the urinary [TETA + MAT] than with nonmetabolized urinary TETA alone. MAT may thus be implicated in the mechanism by which TETA extracts excess systemic Cu in diabetes.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
G.G. is a consultant to, J.L., S.P., and A.A.O. are employed by, and G.C. is a Shareholder and Director of Protemix.
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
ABBREVIATIONS: TETA, triethylenetetramine; ACS, N1-acetylspermine; ADME, absorption, distribution, metabolism, and excretion; DAT, N1,N10-diacetyltriethylenetetramine; HFBA, heptafluorobutyric acid; HPLC, high-pressure liquid chromatography; LC-MS, liquid chromatography-mass spectrometry; MAT, N1-acetyltriethylenetetramine; NAT, N-acetyltransferase; PABA, p-aminobenzoic acid; ICPMS, inductively coupled plasma mass spectrometry; HFBA, heptafluorobutyric acid; SIM, selected-ion monitoring; LME, linear mixed-effects model; LLOQ, lower limit of quantification.
Address correspondence to: Professor Garth J.S. Cooper, School of Biological Sciences, University of Auckland, Private Bag 92019, Auckland, New Zealand. E-mail: g.cooper{at}auckland.ac.nz
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  H.-Y. Cho, R. A. Blum, T. Sunderland, G. J. S. Cooper, and W. J. Jusko Pharmacokinetic and Pharmacodynamic Modeling of a Copper-Selective Chelator (TETA) in Healthy Adults J. Clin. Pharmacol., August 1, 2009; 49(8): 916 - 928. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
